Demobilisation of critical contaminants in four typical waste-to-energy ashes by carbonation

www.elsevier.com/locate/wasman

Waste Management 26 (2006) 430–441

Demobilisation of critical contaminants in four typicalwaste-to-energy ashes by carbonation

Jelena Todorovic *, Holger Ecke 1

The Division of Waste Science and Technology, Lulea University of Technology, S-971 87 Lulea, Sweden

Accepted 18 November 2005Available online 5 January 2006

Abstract

Two bottom ashes, one air pollution control (APC) residue and one fly ash from three different Swedish municipal solid waste incin-eration (MSWI) plants were characterised regarding the leaching of environmentally relevant components. Characterisation was per-formed using a diffusion tank leaching test. The impact of carbonation on the release of eight critical components, i.e., Cl�, Cr, Cu,Mo, Pb, Sb, Se, SO2�

4 and Zn, was assessed at a lab-scale and showed carbonation to have a more pronounced demobilising effecton critical components in bottom ashes than in APC residue and fly ash. From grate type incinerator bottom ash, the release of Crdecreased by 97%, by 63% for Cu and by 45% for Sb. In the investigated APC residue, the releases of Cr, Se and Pb were defined ascritical, although they either remained unaffected or increased after carbonation. Cl� and SO2�

4 remained mobile after carbonation inall investigated residues.� 2005 Elsevier Ltd. All rights reserved.

1. Introduction

The present strategy of the European Union (EU) forwaste management favours incineration over direct landfill-ing, by far the most dominant waste management optionfor decades (Williams, 2005). Treatment of municipal solidwaste (MSW) by incineration provides mass and volumereduction, disinfection, reduction of organic matter andthe possibility of energy recovery. While 20–30 wt% remainas bottom ash, the cleaning of incineration flue gases gen-erates residues equal to 3–5 wt% of incinerated waste.MSW incineration (MSWI) bottom ashes are either land-filled or reused in road construction. The potentially highleaching of salts and other elements from bottom ashescould increase the price of their disposal if limit valuesfor the acceptance of waste to a lower landfill class areexceeded (EC, 2002). If MSWI bottom ash is reused in

0956-053X/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.wasman.2005.11.011

* Corresponding author. Tel.: +46 920 49 3020; fax: +46 920 49 2818.E-mail addresses: [email protected] (J. Todorovic), holger.ecke@

ltu.se (H. Ecke).1 Present address: International Institute for Applied Systems Analysis

(IIASA), A-2361, Laxenburg, Austria.

the construction of roads, this high release could generatetoxic leachates (Ore et al., submitted). Owing to their fineparticle size and high content of persistent organic pollu-tants (POPs), salts and metals, air pollution control(APC) residues and fly ashes have to be treated before finaldisposal. Due to the reasons mentioned above, finding newtreatment techniques that result in the demobilisation ofcritical MSWI residue components are necessary.

Recent research on carbonation (Ecke et al., 2003;Meima et al., 2002; Polettini and Pomi, 2004) and naturalweathering of MSWI residues (Chimenos et al., 2003;Meima and Comans, 1999) demonstrate these proceduresto possibly decrease the release of contaminants, i.e., Cu,Mo, Pb, and Zn. Due to their composition, MSWI residuesoften possess self-binding properties when mixed withwater and compacted (Chandler et al., 1997), i.e., in mostreuse and disposal scenarios, resulting in their solidifica-tion. The mechanism controlling leaching in solidifiedmaterial (often diffusion controlled) differs from the vari-ous shaking tests often used in regulatory (EC, 2002) andresearch purposes. The release of components from thetested residues by the shaking test is either availability orsolubility controlled, with the exclusion of possible solidifi-

mailto:[email protected]

mailto:holger.ecke@

J. Todorovic, H. Ecke / Waste Management 26 (2006) 430–441 431

cation occurring in landfill conditions. In this work, fourincineration residues are characterised before and afterlab-scale carbonation. The evaluation was based on resultsfrom the diffusion leaching test to consider the conditionsoccurring in landfills or upon reuse. The objective was toassess the effect of carbonation on the demobilization ofcritical components from the compacted residues of boththe combustion chamber and air pollution control (APC)units of three Swedish waste-to-energy plants. Criticalcomponents were identified based on the regulatory criteriafor acceptance of waste to landfills, as stipulated by the EU(EC, 2002). Methods for lab-scale treatment and character-isation are developed and evaluated.

2. Material and methods

2.1. Material

Four ashes from three Swedish incineration plants wereinvestigated. Their types, origin and abbreviations are pre-sented in Table 1.

The stocker grate type (SGT) incinerator mostly receivesMSW and smaller fractions of sorted industrial waste, like

Table 1Origins and types of ashes investigated

Abbreviation Origin Type of ash

SGT-BA Stocker grate typeincinerator

Bottom ash

CFB1-APC Circulating fluidisedbed type incinerator 1

Air pollutioncontrol residue

CFB2-BA Circulating fluidisedbed type incinerator 2

Bottom ash

CFB2-FA Circulating fluidisedbed type incinerator 2

Fly ash

Table 2Amount of total solids (TS), loss on ignition (LOI), oxides and elements in SGT(accepted)), CFB2-BA and CFB2-FA

SGT-BA CFB1-APC

Average SD Average SD

Content of total solids (g (kg ash)�1)

TS 830 ± 3 993.6 ± 1.9

Loss on ignition (g (kg TS)�1)

LOI 12 ± 0.8 5.6

Minor constituents (mg (kg TS)�1)

As 33 ± 3.2 309 ± 6Cd 6 ± 1.1 49.7 ± 8.8Co 34 ± 6 34.5 ± 7.0Cr 568 ± 435 835 ± 112

Cu 11,570 ± 3261 7160 ± 1030

Mo 24 ± 2.6 27.0 ± 5.9Ni 567 ± 446 166 ± 36Pb 2260 ± 983 3310 ± 554S 5103 ± 38 33,300

Sb 59.1 326Zn 9117 ± 1198 10,100 ± 1460

Values are presented as averages and standard deviations (SD). Values writChandler et al. (1997).

wood, rubber and plastic. The plant’s magnetic separatorremoves magnetic particles from the bottom ash. Approx-imately 200 kg of bottom ash (SGT-BA) were sampledfrom a 6-months old heap. Of this, 15 kg was furthersub-sampled in the laboratory. Half of the sub-sampledash was stored in an atmosphere of nitrogen gas, whilethe other half was carbonated. Concentrations of Cu, Sand Zn in the SGT-BA were higher than the typical con-centration ranges of bottom ashes presented by Chandleret al. (1997) (Table 2). The leaching of six components,i.e., Cl�, Cr, Cu, Mo, Sb and SO2�

4 (Table 3), from SGT-BA does not meet the EU Council stipulation for accep-tance of waste to landfills for inert waste (EC, 2002).

The circulating fluidised bed type (CFB) incinerator 1 issupplied with crushed wood, paper and plastic waste. Mag-netic separation is performed at the waste feed beforeentering the kiln. The incinerator uses a dry process toclean the flue gases. Activated carbon is added to the gasstream to remove heavy metals, in particular mercury,while lime is added to remove HCl and H2SO4. Approxi-mately 100 kg of APC residue (CFB1-APC) were sampled,of which 6 kg were sub-sampled in the laboratory andstored in nitrogen gas. The rest was stored in closed plasticbuckets for another 13 months, after which 2 kg was sub-sampled for carbonation. The total concentrations of Cr,Cu and S in CFB1-APC were higher than typical concen-trations from the dry APC system residue presented byChandler et al. (1997) (Table 2). The leaching of three com-ponents, i.e., Cl�, Cr and Se (Table 4), from CFB1-APCdoes not meet the EU’s stipulation for acceptance of wasteto landfills for non-hazardous waste (EC, 2002).

The circulating fluidised bed type incinerator 2 is sup-plied with sorted MSW, wood and wood chips. The incin-erator is equipped with an electrostatic filter, fabric filter,

-BA (from Ecke and Aberg (in press)), CFB1-APC (from Todorovic et al.

CFB2-BA CFB2-FA


999.7 ± 4.0 995.5 ± 9.3

0.0 6.6

64.1 ± 0.9 129 ± 42.5 ± 0.5 27.9 ± 4.824.9 ± 5.0 25.2 ± 5.1305 ± 41 487 ± 665670 ± 813 4530 ± 655

<6 22.9 ± 5.250.9 ± 11.1 129 ± 281040 ± 182 2420 ± 408866 27,60063.9 261237 ± 40 5420 ± 787

ten in bold are higher that respective typical value ranges presented by

Table 3The comparison of compliance leaching test results for non-carbonatedSGT-BA presented as average and standard deviations (n = 5) and thelimit values stipulated by The Council of European Union (EC, 2002) foracceptance of waste to landfills for inert waste (Todorovic et al., accepted)

SGT-BA Limit values

L/S 2 L/S 10 L/S 2 L/S 10


pH 11.5 0.1 11.5 0.1Redox (mV) 53.2 11.4 60.7 25.9

Component (mg (kg TS)�1)

Al 123 29 467 46As <0.016 <0.024 0.1 0.5Ba 0.25 0.02 1.08 0.07 7 20Ca 154 6 858 29Cd 0.0022 0.0003 0.03 0.04Chloride 3402 71 4597 95 550 800

Co 0.0016 0.0001 <0.003Cr total 0.38 0.02 0.78 0.03 0.2 0.5

Cu 1.60 0.05 2.70 0.06 0.9 2

Fluoride 4 10Fe <0.01 <0.02Hg <0.00005 <0.00021 0.003 0.01Mg <0.4 <2Mn 0.003 0.001 <0.01Mo 1.06 0.02 1.83 0.06 0.3 0.5

Ni 0.006 0.001 0.012 0.001 0.2 0.4Pb 0.006 0.004 0.08 0.06 0.2 0.5Sb 0.06 0.02 0.33 0.08 0.02 0.06

Se 0.0106 0.0008 0.023 0.001 0.06 0.1Zn 0.06 0.02 0.3 0.1 2 4Sulphate 640 280 1470 360 560 1000

DOC 79 2 150 8 240 500

Redox was measured in unfiltered leachate.

Table 4The comparison of compliance leaching test results for non-carbonatedCFB1-APC presented as average and standard deviations (n = 5) and thelimit values stipulated by The Council of European Union (EC, 2002) foracceptance of waste to landfills for non-hazardous waste (Svensson et al.,2005)

CFB1-APC Limit values

L/S 2 L/S 10 L/S 2 L/S 10


pH 11.6 0.2 10.7 0.07Redox (mV) �684 84 39.4 12.5

Component (mg (kg TS)�1)

Al 0.02 0.01 0.71 0.14As <0.00 <0.59 0.4 2Ba 0.00 0.06 1.25 0.08 30 100Ca 18,200 230 28,800 349Cd 0.10 0.02 0.11 0.02 0.6 1Chloride 44,800 570 60,900 870 10,000 15,000

Co <0.00 <0.00Cr total 0.02 0.00 15.5 0.60 4 10

Cu 1.6 0.6 2.2 0.6 25 50Fe <0.00 <0.00Hg <0.0000 <0.0000 0.05 0.2K 5730 66 7740 102Mg 0.5 0.1 17.1 0.8Mn <0.00 <0.00Mo 1.47 0.05 3.61 0.08 5 10Na 4810 52 6460 81Ni 0.001 0.00 <0.00 5 10Pb 3.2 0.6 3.2 0.6 5 10S 758.8 6.7 3980 51Sb 0.00 0.00 0.01 0.00 0.2 0.7Se 0.27 0.01 0.55 0.01 0.3 0.5

Zn 10.6 0.3 10.7 0.3 25 50Sulphate 2120 50 11,920 160 10,000 20,000DOC 16.7 1.5 31.6 4.4 380 800

Redox was measured in unfiltered leachate.

Peristaltic pump

CO2

Wash-bottlewith water

Bucket withash

Gas

Fig. 1. Experimental set-up used for the carbonation of CFB1-APC,CFB2-BA, and CFB2-FA.

432 J. Todorovic, H. Ecke / Waste Management 26 (2006) 430–441

scrubbers and a de-nitrification catalytic reactor for treat-ment of flue gases. Fly ash from both electrostatic and fab-ric filters (CFB2-FA) and bottom ash (CFB2-BA) weresampled, approximately 15 kg each. In the laboratory, lar-ger metal particles were removed from CFB2-BA using aweak magnetic separator (Mortsell torr. sep. – Sala mask-infabriks AB, Sweden). Half of the CFB2-BA was stored innitrogen gas, while the other half was carbonated. CFB2-FA was not stored in an inert atmosphere, but the experi-ments started immediately.

Sub-samples of all four ashes were used for the diffusionand availability leaching tests, as well as to analyse chemi-cal composition and content of total solids. Sub-samples ofSGT-BA, CFB1-APC and CFB2-BA (all stored in the inertatmosphere), as well as CFB2-FA (tested without carbon-ation), are referred to in following text as fresh or non-car-bonated samples.

2.2. Carbonation

SGT-BA was carbonated in 50-l gasbag equipped with agas-vent (Tecobag for gas-analyses, Tesseraux, Germany).After the bottom ash was placed in the bag, the air wasremoved through the gas-vent and pure CO2 was added.No water was added. Bottom ash was stored in an atmo-

sphere of CO2 for 27 days. The gasbag was manually sha-ken every other day to mix the bottom ash inside the bag.

CFB1-APC was placed in plastic buckets and CO2 waspumped through the layer of ash (Fig. 1). The flow ofCO2 was set at approximately 33 l (day · kg ash)�1.CFB1-APC was carbonated for 25 days.

CFB2-BA was stored in a gasbag (as described for SGT-BA) for 12 months. Since no significant pH decrease was


observed, carbonation was continued in plastic buckets(Fig. 1) for another 49 days.

CFB2-FA was carbonated in plastic buckets (Fig. 1) for49 days.

2.3. Preparation of specimens for the diffusion leaching test

Specimens were comprised of non-carbonated and car-bonated samples from each ash. Three specimens weremade of non-carbonated SGT-BA and four specimens ofeach other ash.

The ash was mixed with deionised water (Table 5), with-out the addition of any other binding agents. The mixturewas then compacted in layers, using a procedure similar toProctor compaction (SS, 1994). Plastic laboratory beakers(10.5 cm in diameter) or plastic tube-moulds (5 cm in diam-eter and 10 cm high) were used to mould the mixtures(Table 5). After compaction, the specimens were left tocure (Table 5), followed by one specimen from each groupbeing stored. Stored specimens were later used for avail-ability testing, while the others were subjected to a diffusionleaching test.

2.4. Diffusion leaching test

Based on the Dutch diffusion leaching test NEN 7345(NEN, 1995), the diffusion leaching test was carried outin eight successive leaching steps of specified lengths(0.25, 1, 2.25, 4, 9, 16, 32 and 64 days), yielding eight leach-ing fractions from each specimen. Distilled water acidifiedto pH 4 was used as leachant. The only modification fromthe original test protocol was that the specimens were notde-moulded after drying. The leaching of SGT-BA speci-mens was performed in beakers used as moulds (Fig. 2),

Table 5

Conditions for specimen preparation and their dimensions

Ash Compaction

and

curing

Addition of

deionised

water (l kg�1)

Curing

temperature

(�C)

Specimen

dimensions

Diameter

(cm)

Height

(cm)

SGT-BA In beaker 0.1 60 10.5 5–5.5

CFB1-APC In mould 0.2 20 5 9–9.5

CFB2-BA In mould 0.1 20 5 9–9.5

CFB2-FA In mould 0.2 20 5 9–9.5

All specimens were cured for 15 days.

Fig. 2. Experimental set-up for the diffusion leaching test for: (a) SGT-BAand (b) CFB1-APC, CFB2-BA, and CFB2-FA.

with only the top surface (10.5 cm in diameter) beingexposed to the leachant. Specimens prepared in tube-moulds were placed in beakers, where both specimen bases(5 cm in diameter each) were exposed to leachant and theremainder of the surface was covered by the tube mould(Fig. 2). Diffusion leaching test results were presentedand evaluated as cumulative release after four steps forZn and after eight steps for all other investigatedcomponents.

2.5. Availability test

The availability of components was tested according tothe Nordtest method NT Enviro 003 (Nordtest, 1995), con-ducted on material ground to a particle size of 95 wt%<125 lm. A Coulter Multisizer II (Coulter ElectronicsLimited, Luton, UK) determined the particle size by mea-suring the particle volume via the use of an electrical sens-ing zone.

The two-step availability leaching test was carried out ata liquid-to-solid (L/S) ratio of 100 l (kg TS)�1 each. Thefirst step was performed at pH 7 for 3 h, and the secondat pH 4 for 18 h. A computer-controlled automatic titrator(ABU, 901, Radiometer, Copenhagen, Denmark) main-tained the pH during the leaching steps. Both leachate frac-tions were combined prior to analyses.

The amounts of components leached out during theavailability test will now be referred to as amounts avail-able for leaching or availabilities.

2.6. Analyses and analytical methods

The pH was determined using a pH meter pH 340(WTW, Wielheim, Germany). Leachates were analysedfor elements, chloride and sulphates. Elements in theSGT-BA leachate were determined using modified EPAmethods 200.7 (ICP-AES) and 200.8 (ICP-SMS). Elementsin all other leachates were determined using methods SS-EN ISO 11885 and EPA 6020. Chlorides were determinedby titration, according to Swedish standards SS 02 81 20(SS, 1974). Sulphates were determined using SS-EN ISO10304-1. The content of total solids (TS) in ashes was deter-mined according to Swedish standard SS 02 81 13 (SS,1981).

3. Results and discussion

3.1. Assessment of methods

A drop in ash pH (Fig. 3) indicated the occurrence ofcarbonation. Ashes were carbonated dry in two differentways, i.e., by storing the ash in an atmosphere of CO2

and by pumping CO2 through the layer of ash. The firstcarbonation procedure resulted in a pH significant dropin only the case of SGT-BA. The water content of SGT-BA (Table 2) probably explains the occurrence of carbon-ation and pH drop in this ash, since the presence of water

11.25

12.45 12.51

8.28

9.77 10.04

12.1411.42

7

9

11

13

SGT-BA CFB1-APC CFB2-BA CFB2-FA

pH

non-carbonated carbonated

Fig. 3. pH values of ashes before and after carbonation measured at L/S10 l (kg ash)�1. SGT-BA was carbonated in the CO2 atmosphere for 27days, CFB2-BA was stored in a CO2 atmosphere for 12 months andcarbonated with pumped CO2 (Fig. 1) for 49 days, while CFB1-APC wascarbonated with pumped CO2 (Fig. 1) for 25 and CFB2-FA for 49 days.


enhances carbonation (Bergman, 1996). The other threeashes were carbonated by pumping CO2 through the layerof ash (Fig. 1). Moisture was provided to the systemthrough saturation of the gas with water. Continuouspumping of CO2 through the layer of ash was assumedto provide constant contact between the fine ash particlesand CO2.

To keep laboratory procedures on time and resourceefficient, the carbonations of CFB2-BA and CFB2-FAwere stopped after 49 days, even though their pH did notdecrease as much as SGT-BA and CFB1-APC (Fig. 3). Itwas assumed that after 49 days of continuous contact withCO2, some ash stabilisation might have occurred. Ecke(2001) stresses that stabilisation through carbonation isnot only due to changes in pH, but also due to chemicalredistribution. Marked changes in element leaching fromshort-term weathered MSWI bottom ash were observedby Chimenos et al. (2003), even though a significant changein pH did not occur.

The diffusion leaching test was chosen for this investiga-tion because it is one of the few standardised tests for com-

6

8

10

12

0.1 1 10 100days

pH

7

9

11

13

0.1 1 10 100days

pH

SGT-BA

CFB2-BA

Fig. 4. pH development during diffusion leaching test. Average pH values of leash (closed circles) with error bars representing standard deviations (n = 2 for

pacted and monolithic materials. Materials were testedcompacted to simulate field conditions when ashes arelandfilled or reused as secondary construction materials.A design without the addition of binding agents was chosento evaluate the effect of only carbonation on the leaching ofcritical components. Ashes were carbonated dry to avoidinitiating pozzolanic and cementitious reactions during car-bonation. Instead, these reactions were initiated when ash/water mixtures were prepared, right before compaction.This permits evaluating the effects of carbonation andsolidification separately (upon mixing with water and com-paction) (Todorovic et al., accepted). It was also expectedthat mechanically a more stabile and solid specimen wouldbe produced.

Mechanisms controlling mass transport between a solidsurface and water are rather pH-dependent (Stumm, 1992).Changes in pH during leaching could thus affect the releaseof many environmentally relevant components, e.g., Pb, Cuand elements forming oxyanions. Carbonation decreasedthe pH of ashes (Fig. 3), but only the pH of carbonatedSGT-BA reached the pH-value of calcite equilibrium withatmospheric CO2 (pH � 8.3). During the entire diffusiontest, only the pH values of leachate fractions from carbon-ated SGT-BA remained lower than those of the non-car-bonated. All other ashes exhibited similar pH values ofleachates coming from treated and untreated ashes in laterleaching steps (Fig. 4).

Although acidic (pH 4) leachant was used for the diffu-sion tests, highly alkaline ashes were expected to produceleachates of pH values similar to those measured in theash suspensions (Fig. 3) during the last and the longest stepof the test (Fig. 4). This was the case for the carbonatedand non-carbonated samples of the two tested bottomashes and the carbonated sample of CFB1-APC. Fromthe non-carbonated CFB1-APC and both CFB2-FA sam-

7

9

11

13

0.1 1 10 100days

pH

7

9

11

13

0.1 1 10 100days

pH

CFB1-APC

CFB2-FA

achate fraction from non-carbonated ash (open triangles) and carbonatednon-carbonated SGT-BA and n = 3 for all other ashes).


ples, the final leachates had considerably lower pH-valuesthan those measured in ash suspensions. The lower pH ofdiffusion test fractions could possibly be due to ash becom-ing carbonated after measuring the pH in ash suspensions,i.e., pH of ashes was lowered even further to the valuesbelow those noted in Fig. 3 during the preparation of solidspecimens. Carbonation was probably enhanced by waterprovided before compaction (Table 5), while air providedCO2 during mixing of the ash–water mixtures and curingof samples. Because beakers used for the diffusion test werenot airtight, it was expected that some carbonation wouldoccur even during leaching.

3.2. Critical components

The compositions of CFB2-BA and CFB2-FA are pre-sented in Table 2. Due to their high content (Table 2) orhigh release (Tables 3 and 4) or both, the leaching of thefollowing components was chosen to be discussed here:Cl�, Cr, Cu, Mo, Pb, Sb, Se, SO2�

4 and Zn. A high contentof Cr was observed in CFB1-APC, of Cu in SGT-BA,CFB1-APC and CFB2-FA, of S in SGT-BA and CFB1-APC, and of Zn in SGT-BA (Table 2). Releases of Cl�,Cr, Cu, Mo, Sb, Se and SO2�

4 were found to be critical fromSGT-BA or CFB1-APC or both, based on the complianceleaching test (Tables 3 and 4). The release of Cl� fromCFB1-APC determined during the compliance leaching test(60.9 ± 0.9 g (kg TS)�1) was not consistent with the releaseof Cl� assessed during the availability test (47.5 ± 0.1 g(kg TS)�1), since the latter is considered to be the entireleachable amount. This inconsistency could be due to themethod used to determine Cl� in water samples (SS,1974) being based on the titration with AgNO3. Interfer-ences in more concentrated leachates from the compliancetest could mask the colour change and show a higher con-tent of Cl�. Leaching of Pb from CFB1-APC at L/S 2 waslower than the limit value, but still relatively high. Both Pband Zn form carbonates or could be trapped by secondaryminerals (Piantone et al., 2004). Previous work on MSWIresidues (Ecke, 2001; Todorovic et al., accepted) shows thattheir release could be significantly affected by the carbon-ation process, which thus supports their inclusion in thisstudy.

The amounts of critical components available for leach-ing are shown in Fig. 5. The availability of all investigatedcomponents was lowest from CFB2-BA. Carbonationincreased the availability of Pb by 35% and Zn by 18% inSGT-BA, of Cu by 127% in CFB1-APC, and of Sb by81% and Zn by 18% in CFB2-FA. In SGT-BA, however,the availability of Cr decreased by 45%, Mo by 37% andSb by 62%, as well as that of Cr by 28%, Pb by 12%and Se by 18% in CFB1-APC, Pb by 72% in CFB2-BAand Se by 44% in CFB2-FA. Carbonation decreased thecumulative release of Cr, Mo and Sb from both bottomashes (Fig. 6). From CFB2-FA, the cumulative release ofSe decreased due to carbonation, while the cumulativerelease of Mo and Pb increased. Carbonation also

increased the cumulative release of Cu from CFB1-APCand CFB2-BA, but decreased that from SGT-BA.

3.3. Anions

The leaching of Cl� is independent on pH, relativelyrapid and not controlled by solubility limitations (Slootet al., 1997). An increase in cumulative leaching of Cl�

from SGT-BA and CFB1-APC after carbonation (Fig. 6)is thus unexpected. An explanation for such results couldbe similar to that describing the inconsistency of Cl� leach-ing results in the paragraph Critical components. Recentfindings (Todorovic et al., accepted) show it unlikely thatchanges in leaching from SGT-BA are caused by changesin physical retention after carbonation. From CFB2-BAand CFB2-FA, the Cl� leaching from carbonated andnon-carbonated ash was within the standard deviation.Since the leaching of Cl� could not be decreased by car-bonation, the most efficient solution seems to be todecrease the content of Cl� by washing the ash. The wash-ing process to remove salts from MSWI ashes has alreadybeen investigated and applied (Chandler et al., 1997).

The highest release of SO2�4 was observed from CFB1-

APC (Figs. 5 and 6), probably due to the high concentra-tion of relatively soluble CaSO4 in the ash. This CaSO4

in the APC residue originates from SO2 that was cleanedfrom the flue gas stream through the addition of Ca(OH)2.Furthermore, carbonation increased the cumulative releaseof SO2�

4 from SGT-BA by 320% and by 63% for CFB1-APC. An increase in SO2�

4 leaching from aged and carbon-ated MSWI bottom ash was also observed by Steketee andUrlings (1994) and Meima et al. (2002). Ash mineralogywas not investigated in this study, but numerous studies(Meima and Comans, 1997; Polettini and Pomi, 2004;Zevenbergen and Comans, 1994) show that ettringite(Ca6Al2(SO4)3(OH)12 · 12H2O) could be present in bottomash that is freshly quenched, several weeks old and weath-ered. Carbonation, however, decomposes ettringite (Xian-tuo et al., 1994). It could be expected that in carbonatedSGT-BA with a pH 8.28, SO2�

4 ions from ettringite wouldbe regrouped to more soluble species. This process offersone explanation for the increase in SO2�

4 leaching fromSGT-BA due to carbonation. Contrary to SGT-BA andCFB1-APC, the cumulative release of SO2�

4 from CFB2-BA and CFB2-FA (Fig. 6) was low and unaffected by car-bonation. After 1 day of leaching, the concentrations ofSO2�

4 in leachates from CFB2-BA were close to or belowthe detection limit.

3.4. Metals

The work of Ecke et al. (2003) onMSWI fly ash shows Pband Cu to be demobilised by two orders of magnitude aftercarbonation. Numerous investigations on artificially car-bonated and naturally weathered bottom ashes (Chimenoset al., 2003; Meima and Comans, 1999) also show a decreasein Pb and a Zn release. The increases in Pb cumulative

0

20

40

60

80

100

Non-carbonated

Carbonated

Cr

0

1000

2000

3000 Cu

0

2

4

6

8

10

12 Mo

0

1000

2000

3000

4000 Zn

0

20

40

60Sb

0

1

2

3

4Se

0

500

1000

1500

2000 Pb

Ava

ilabl

e (m

g/(k

g T

S))

0

20000

40000

60000

80000

SO4

0

10000

20000

30000

40000

50000

60000

70000










Cl

Fig. 5. Availability of components from ashes before and after carbonation.


leaching from SGT-BA (by 159%) and CFB2-FA (by 64%)due to carbonation as well as Zn from SGT-BA (almost7-fold over 4 days) were therefore unexpected.

The pH of leachates from the first four steps of the dif-fusion test (that Zn was analysed for) from carbonatedSGT-BA ranged between 6.98 and 7.19, while from non-carbonated SGT-BA was between 7.42 and 9.21. V-shapedleaching vs. pH curve for Zn have their minimum pHbetween 9 and 11 (Chandler et al., 1997; Sloot et al.,1996). The increase in Zn leaching after carbonation couldthus be explained by a higher solubility of amphoteric Znat pH range of carbonated SGT-BA. A similar explanationcould be applied to the leaching of Pb from SGT-BA, since

Pb exhibits its lowest solubility at pH between 9 and 10(Chandler et al., 1997).

The release of Pb from CFB1-APC during complianceleaching test (Table 4) occurred predominantly during thefirst step (at L/S 2 l (kg TS)�1), amounting to3.2 ± 0.6 mg (kg TS)�1. In CFB1-APC, Pb occurs in theform of soluble hydroxides (Fig. 7), which explains thisrapid release. The cumulative release from non-carbonatedCFB1-APC after eight steps of the diffusion test (Fig. 6)was 0.041 ± 0.005 mg (kg TS)�1. One reason for thismarked decrease compared to compliance leaching test,could be the demobilisation of Pb by a process of solidifi-cation. Here, recognizing Pb as critical could also be ques-

Fig. 6. Cumulative release of Cl, Cr, Cu, Mo, Pb, S, Sb, and Se after 8 steps (64 days) of diffusion test leaching, and of Zn after 4 steps (4 days).


tioned. Fly ashes and APC residues are either landfilled orreused in a solidified form, or are expected to solidify uponcontact with water in the field where leaching conditionscorrespond to those simulated by the diffusion test. Thecompliance leaching (shaking) test as stipulated by theEuropean Council (EC, 2002) as a criterion for acceptanceof waste to landfills obviously overestimated the release ofPb from fly ashes and APC residues.

While carbonation did not significantly impact thecumulative release of Pb from CFB1-APC, that from

CFB2-FA increased by 64% after carbonation (Fig. 6).Because each diffusion test step is longer than the previous,it could be expected that both pH and elemental concentra-tion in the resulting leachate would increase from one stepto another. This is confirmed for non-carbonated CFB2-FA, but not for the carbonated (Fig. 8). The most signifi-cant difference was observed during the first three stepsof the diffusion leaching test, when Pb was washed-outfrom carbonated ash, indicating that the carbonation ofCFB2-FA led to a transition of Pb to easily soluble forms.

7 9 11 13

pH

-6

0

6

pe

Pb3(

OH

) 42+

Pb(O

H) 4

2-

Pb(O

H) 3

–

PbCl +

PbO

H+

Fig. 7. Predominance area diagram for Pb phases in CFB1-APCaccording to calculations using PHREEQC-2 (Parkhust and Appelo,2004) from Svensson et al. (2005).

0

0.02

0.04

0.06

8 9 10 11 12

pH

Pb

(mg/

(kg

TS)

)

Fig. 8. Correlation between the release of Pb and end-point pH duringdifferent steps of diffusion leaching of CFB2-FA. Black marks show therelease from carbonated samples, while the white ones show the non-carbonated ones.er – step 1,nm – step 2,hj – step 3,sd – step 4,er – step 5, n m – step 6, h j – step 7, and s d – step 8.

0.00

0.04

0.08

0.12

1 2 3 4 5 6 7 8

step

Cu

(mg/

kg T

S)

7.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

pH

Fig. 9. Release of Cu from CFB1-APC during the diffusion leaching test(bars) compared with pH values of resulting leachates (squares connectedwith lines), all presented as averages (n = 3). Black presents values forcarbonated sample, white for non-carbonated one. Note that time lengthincreases from one step to another.

6

7 9 11 13

pH

-6

0pe Cr(OH)3

CrO2-

CrO42–

(s)

7 9 11 13

pH

-6

0

6

pe

Cr(OH)3CrO2

-

CrO42–

Cr(OH)2+

(s)

(a) (b)

Fig. 10. Predominance area diagram for Cr phases in SGT-BA (a) andCFB1-APC (b) according to calculations using PHREEQC-2 (Parkhustand Appelo, 2004) from Svensson et al. (2005).


In later stages of the test, both the Pb release and pH valueswere similar.

Carbonation decreased the cumulative release of Cufrom SGT-BA to almost one-third (Fig. 6). Most Cu fromMSWI bottom ash of different ages is mobilised by dis-solved organic carbon (DOC) (Meima et al., 1999). Recentfindings (Todorovic et al., accepted) demonstrated artificialcarbonation of suspended SGT-BA with an excess of car-bon-dioxide (pH 6.40 ± 0.07 (n = 8)) to increase both therelease of Cu and DOC in leachate. This is one reasonwhy carbonation to a moderate alkaline pH or naturalweathering should be preferred when the treatment fordemobilisation of Cu in MSWI bottom ash is considered.Another reason is that in naturally weathered bottomash, sorption to (hydr)oxides of Al and Fe plays an impor-tant role in demobilising Cu (Meima and Comans, 1999). Itcould also be expected that more sorption places aregradually opened, further decreasing the release of Cu.Carbonation increased the cumulative release of Cu in bothCFB1-APC (by 3 times) and CFB2-BA (by 2 times). Theavailability of Cu (Fig. 5) from CFB1-APC also increased

after carbonation. Fig. 9 shows that during leaching ofnon-carbonated CFB1-APC, the concentrations of Curemained relatively constant and unaffected by time ofleaching or change in pH. However, the release of Cu fromcarbonated ash initially increases with the leaching timeand pH and then decreases in later steps. With such inten-sive leaching from stabilised carbonated material, there is arisk of increased Cu leaching during compliance testing(run on granular material without pH adjustment). Careshould be taken to not exceed the limit values for accep-tance of waste to landfills for non-hazardous waste (EC,2002) after carbonation of CFB1-APC. The release of Cufrom both CFB2-BA (although it increased after carbon-ation) and CFB2-FA are not considered critical.

3.5. Elements forming oxyanions

In natural systems, Cr occurs as a relatively stable triva-lent (Cr(III)) and highly toxic and mobile hexavalent(Cr(VI)). At a pH of non-carbonated SGT-BA (pH11.25) and oxidising conditions, most Cr is expected tobe present as Cr(VI) (Fig. 10(a)). The reduced Cr leaching(by 97%) was probably due to the decreasing pH to thevalue of carbonated SGT-BA (pH 8.28), since the predom-


inant compound at moderately alkaline conditions is insol-uble Cr(OH)3. Only 1.3 wt% of Cr (8.83 ± 1.08 mg(kg TS)�1) was available for leaching in SGT-BA (Fig. 5).This low leaching also proves that most Cr probably pres-ent in SGT-BA was in a trivalent oxidation state.

The cumulative release of Cr from CFB1-APC, highestcompared to other investigated ashes, increased after car-bonation by 60% (Fig. 6). However, this cumulative releasewas two orders of magnitude lower than both the availabil-ity of Cr in CFB1-APC (Fig. 5) and its release during thesecond step of the compliance test (Table 4). The pH ofall diffusion test leachate fractions (Fig. 4) ranges between7.3 and 10.5, where insoluble Cr(OH)3 is the dominatingcompound below pe 3 (Fig. 10(b)). Increased Cr leachingduring the second step of the compliance leaching test(Table 4), when oxidising conditions prevailed, proves thatachieving reducing conditions might be important formaintaining a low Cr release. These conditions might notbe difficult to achieve, since the formation of hydrogen ispossible in flue-gas cleaning residues from the incinerationof waste in the fluidised bed. Hydrogen is a product of Alhydrolysis that at a high pH could simultaneously reduceCr(VI) to Cr(III) (Abbas et al., 2001). When consideringcarbonation to treat MSWI residues, carbonation with agas of high CO2 content might be a preferable method overa treatment with air. This could reduce the contact of thetreated residue with oxygen and maintain reducing condi-tions for the demobilisation of Cr.

The relatively high cumulative release of Se from CFB1-APC (0.59 ± 0.05 mg (kg TS)�1) and CFB2-FA(0.71 ± 0.06 mg (kg TS)�1) remained unaffected by carbon-ation. In natural systems, Se occurs in a number of valentstates (+6, +4, +2, and as elemental) (Sawyer et al., 2003).Reducing Se could demobilise this element, since Se(IV)could be better retained than Se(VI) (Hoek and Comans,1994). Chemical equilibrium calculations (Fig. 11) showedthe occurrence of Se as Se(IV) in CFB1-APC when reduc-ing conditions were achieved in highly alkaline conditions.Contrary to Cr, Se showed a relatively high release(0.27 ± 0.01 mg (kg TS)�1) under reducing conditions ofthe first step of the compliance leaching test (Table 4).

pe

7 9 11

pH

-6

0

6

HSe-

SeO32-

SeO42–

13

HSe

O3-

MnSe

Cd(

SeO

3)22–

(s)

Fig. 11. Predominance area diagram for Se phases in CFB1-APCaccording to calculations using PHREEQC-2 (Parkhust and Appelo,2004) from Svensson et al. (2005).

The cumulative release of Sb during the diffusion leach-ing of both bottom ashes decreased due to carbonation (inSGT-BA by 45% and in CFB2-BA by 26%). Recent find-ings (Todorovic et al., accepted) show the negative loga-rithm of the mean effective diffusion coefficients (pDe) toslightly decrease with carbonation (from 12.2 ± 0.0 fornon-carbonated to 12.0 ± 0.0 for carbonated sample), i.e.,a faster release of Sb from compacted carbonated SGT-BA than from non-carbonated. Hence, the decrease of Sbleaching from SGT-BA was probably due to a decreasein its availability (Fig. 5). As for Cu, the same investiga-tions (Todorovic et al., accepted) also show an increasein Sb leaching from SGT-BA after excessive carbonation(pH 6.40 ± 0.07 (n = 8)). Meima and Comans (1998) sug-gest that for demobilisation of Sb, more slightly alkalineconditions are preferred, such as those achieved by car-bonation of SGT-BA presented in this study. At slightlyalkaline values, Sb shows an affinity for sorption to Al-and Fe-(hydr)oxides (Meima and Comans, 1998). As inthe case of Se, the cumulative release of Sb from flue gascleaning residues remained unaffected by carbonation.

The difference between the cumulative release of Mofrom non-carbonated and carbonated SGT-BA was withina relatively high standard deviation (Fig. 6). If one leachatefraction with an extremely high Mo concentration wasexcluded from the analyses, the cumulative release of Mofrom carbonated SGT-BA was lower (32%) than fromnon-carbonated SGT-BA. As with Sb, the decrease in thecumulative release of Mo is due to its decrease in availabil-ity, since its release from carbonated and non-carbonatedcompacted SGT-BA should be equally rapid (Todorovicet al., accepted).

Carbonation also decreased the cumulative release ofMo from CFB2-BA by 50%. As for Sb in weathered MSWIbottom ashes, sorption could be the mechanism controllingthe release of Mo (Meima and Comans, 1999). In CFB1-APC, Mo was not recognised as critical, though carbon-ation mobilised Mo in both flue gas cleaning residues (inCFB1-APC increase was by 50% and in CFB2-FA by41%). When considering carbonation as a treatment ofgas cleaning residues prior to landfilling, care should thusbe taken not to increase the release of Mo over the limitvalues for acceptance of waste to corresponding landfillclass (EC, 2002).

4. Conclusions

The effect of carbonation on the release of critical com-ponents was analysed on four municipal solid waste incin-eration (MSWI) residues. Bottom ash from a stocker gratetype incinerator (SGT-BA), air pollution control residue(CFB1-APC), bottom ash (CFB2-BA) and fly ash (CFB2-FA) from two circulating fluidised bed type incineratorswere all included in the analyses. Components found tobe critical were Cl�, Cr, Cu, Mo, Pb, Sb, Se, SO2�

4 andZn. The leaching of critical components was assessed onashes before and after carbonation using a 64-day diffusion


leaching test. Ashes were carbonated using CO2 withoutthe addition of water.

Carbonation proved to be an ineffective method for thedemobilisation of Cl� and SO2�

4 . The cumulative release ofSO2�

4 increased in one investigated MSWI bottom ash (byfour times in SGT-BA), probably due to the decompositionof ettringite in MSWI residues. The cumulative release ofSO2�

4 also increased in CFB1-APC by 62%.Carbonation was more effective in demobilising other

critical components in bottom ashes (SGT-BA andCFB2-BA) than in the two investigated flue gas cleaningresidues (CFB1-APC and CFB2-FA). Cr, Cu and Sb areall recognised as critical in SGT-BA and were demobiliseddue to carbonation by 96% for Cr, 63% for Cu, and 45%for Sb. The cumulative release of elements forming oxya-nions from CFB2-BA decreased (Cr by 22%, Mo by 50%and Sb by 26%), while the release of Cu increased, butremained low (0.06 mg (kg TS)�1 over 64 days leaching).Sorption probably plays an important role in the demobi-lisation of Cu, Sb and Mo. The release of Pb and Zn wasnot considered critical from the two investigated bottomashes, though they did increase after carbonation, probablydue to a decrease in pH to the values of a higher Pb and Znsolubility. When carbonation is considered to treat MSWIbottom ashes, care should be taken to not cause the releaseof Pb and Zn to increase over stipulated limit values.

Carbonation made less of a promising effect on thedemobilisation of critical constituents Cl�, Cr, Se and Pbfrom the two investigated flue gas cleaning residues(CFB1-APC and CFB2-FA). The cumulative release ofall four constituents either remained unaffected orincreased after carbonation. A possible improvement ofthe applied carbonation procedure could be to treat wetash with gas of a high CO2 content. Reducing conditionscould contribute to a demobilisation of Cr and Se in theseresidues. Using gas with a high CO2 content instead of airfor carbonation could help achieve and maintain thesereducing conditions. Pb is expected to occur in mobileforms in flue gas cleaning residues.

Acknowledgements

The authors thank the Swedish Energy Agency (projectP13099-1), The Swedish Research Council for Environ-ment, Agricultural Science and Spatial Planning (Formas)(project 25.0/2001-0446), and Varmeforsk (project Q4-140) for their financial support. We also thank Henrik Bri-stav (Umea Energi), Maria Nyholm (RagnSells) and MariaVamling (Sundsvall Energi) for providing useful informa-tion and supporting our work.

References

Abbas, Z., Steenari, B.-M., Lindqvist, O., 2001. A study of Cr(VI) in ashesfrom fluidized bed combustion of municipal solid waste: leaching,secondary reactions and the applicability of some speciation methods.Waste Management 21 (8), 725–739.

Bergman, A., 1996. Characterisation of industrial wastes. Licentiate thesis,ISRN HLU-TH-L-1996/21-L-SE. Department of EnvironmentalPlanning and Design, Lulea University of Technology, Lulea, Sweden.

Chandler, A.J., Eighmy, T.T., Hartlen, J., Hjelmar, O., Kosson, D.S.,Sawell, S.E., Sloot, H.A.v.d., Vehlow, J., 1997. Municipal Solid WasteIncinerator Residues. Elsevier Science B.V., Amsterdam.

Chimenos, J.M., Fernandez, A.I., Miralles, L., Segarra, M., Espiell, F.,2003. Short-term natural weathering of MSWI bottom ash as afunction of particle size. Waste Management 23 (10), 887–895.

EC, 2002. Council decision establishing criteria and procedures for theacceptance of waste at landfills pursuant o Article 16 and Annex II ofDirective 1999/31/EC. 14473/02 ENV 682, Council of the EuropeanUnion, Brussels.

Ecke, H., 2001. Carbonation for Fixation of Metals in Municipal SolidWaste Incineration (MSWI) Fly Ash, Doctoral Thesis, LTU-DT-01/33-SE. Department of Environmental Engineering, Lulea University ofTechnology, Lulea, Sweden.

Ecke, H., Aberg, A., in press. Quantification of the effects of environ-mental leaching factors on emissions from bottom ash in roadconstruction. The Science of the Total Environment.

Ecke, H., Menad, N., Lagerkvist, A., 2003. Carbonation of municipalsolid waste incineration fly ash and the impact on metal mobility.Journal of Environmental Engineering 129 (5), 435–440.

Hoek, E.E.v.d., Comans, R.N.J., 1994. Speciation of As and Se duringleaching of fly ash. In: Environmental Aspects of Construction withWaste Materials. Elsevier Science, Amsterdam, The Netherlands, pp.467–476.

Meima, J.A., Comans, R.N.J., 1997. Geochemical modelling of weath-ering reactions in municipal solid waste incineration bottom ash.Environmental Science and Technology 31 (5), 1269–1276.

Meima, J.A., Comans, R.N.J., 1998. Reducing Sb-leaching frommunicipal solid waste incinerator bottom ash by addition of sorbentminerals. Journal of Geochemical Exploration 62 (1–3), 299–304.

Meima, J.A., Comans, R.N.J., 1999. The leaching of trace elements frommunicipal solid waste incinerator bottom ash at different stages ofweathering. Applied Geochemistry 14 (2), 159–171.

Meima, J.A., van Zomeren, A., Comans, R.N.J., 1999. Complexation ofCu with dissolved organic carbon in municipal solid waste incineratorbottom ash leachates. Environmental Science and Technology 33 (9),1424–1429.

Meima, J.A., Weijden, R.D.v.d., Eighmy, T.T., Comans, R.N.J., 2002.Carbonation processes in municipal solid waste incinerator bottom ashand their effect on the leaching of copper and molybdenum. AppliedGeochemistry 17 (12), 1503–1513.

NEN, 1995. Leaching characteristics of solid earthy and stony buildingand waste materials. Leaching tests. Determination of the leaching ofinorganic components from buildings and monolithic waste materialswith the diffusion test. NEN 7345, Nederlands normalisatie instituut,Delft, The Nederlands.

Nordtest, 1995. Solid waste, granular inorganic material: Availability test,NT ENVIR 003. 1096-93, Nordtest, Espoo, Finland.

Ore, S., Todorovic, J., Ecke, H., Grennberg, K., Lagerkvist, A.,submitted. Toxicity of leachate from bottom ash used in a roadconstruction. Waste Management.

Parkhust, D.L., Appelo, A.J., 2004. PHREEQC for Windows, Ahydrochemical transport model. Denver, Colorado, USA.

Piantone, P., Bodenan, F., Chatelet-Snidaro, L., 2004. Mineralogicalstudy of secondary mineral phases from weathered MSWI bottom ash:implications for the modelling and trapping of heavy metals. AppliedGeochemistry 19 (12), 1891–1904.

Polettini, A., Pomi, R., 2004. The leaching behavior of incinerator bottomash as affected by accelerated ageing. Journal of Hazardous Materials113 (1–3), 209–215.

Sawyer, C.N., McCarty, P.L., Parkin, G.F., 2003. Chemistry forEnvironmental Engineering and Science. McGraw-Hill, New York,USA.

Sloot, H.A.v.d., Comans, R.N.J., Hjelmar, O., 1996. Similarities in theleaching behaviour of trace contaminants from waste, stabilized waste,


construction materials and soils. The Science of the Total Environment178 (1–3), 111–126.

Sloot, H.A.v.d., Heasman, L., Quevauviller, P., 1997Harmonization ofLeaching/Extraction Tests, vol. 70. Elsevier Science B.V., Amsterdam,The Nederlands.

SS, 1974. Bestamning av kloridhalt hos vatten, SS 02 81 20. Sverigesstandardiseringskommission, Stockholm, Sweden.

SS, 1981. Vattenundersokningar – bestamning av torrsubstans ochglodgningsrest i vatten, slam och sediment, SS 02 81 13. Sverigesstandardiseringskommission, Stockholm, Sweden.

SS, 1994. Geotekniska provningsmetoder, Packningsegenskaper, Labora-toriepackning, SS 02 71 09. Sveriges standardiseringskommission,Stockholm, Sweden.

Steketee, J.J., Urlings, L.G.C.M., 1994. Enhanced natural stabilisation ofMSW bottom ash: a method for minimisation of leaching. In:Environmental Aspects of Construction with Waste Materials. ElsevierScience, Amsterdam, The Netherlands, pp. 233–238.

Stumm, W., 1992. Chemistry of the Solid–Water Interface. Wiley, NewYork, USA.

Svensson, M., Herrmann, I., Ecke, H., Sjoblom, R., 2005. Selektivmobilisering av kritiska element hos energiaskor (Selective mobiliza-tion of critical elements in incineration ashes). Report number: Q4-104,Varmeforsk Service AB, Stockholm, Sweden.

Todorovic, J., Svensson, M., Herrmann, I., Ecke, H., accepted. Artificialcarbonation for controlling the mobility of critical elements in bottomash. Journal of Material Cycles and Waste Management.

Williams, P.T., 2005. Waste Treatment and Disposal. Wiley, Chichester,England.

Xiantuo, C., Ruizhen, Z., Xiaorong, C., 1994. Kinetic study of ettringitecarbonation reaction.Cement andConcreteResearch 24 (7), 1383–1389.

Zevenbergen, C., Comans, R.N.J., 1994. Geochemical factors controllingthe mobilization of major elements during weathering ofMSWI bottomash. In: Environmental Aspects of Construction with Waste Materials.Elsevier Science, Amsterdam, The Netherlands, pp. 179–194.

Demobilisation of critical contaminants in four typical waste-to-energy ashes by carbonation

Documents

Transcript of Demobilisation of critical contaminants in four typical waste-to-energy ashes by carbonation